Nonparametric or “distribution-free” statistics are a set of statistical methods that make limited assumptions about the population distribution from which the analyzed sample is drawn. These nonparametric statistics contrast with traditional parametric methods that assume the underlying population is characterized by the normal distribution and defined by population parameters (e.g., mean, standard deviation). The nonparametric family of statistics includes both descriptive and inferential statistics, such as the Mann-Whitney U, Spearman’s rho, and Chi-square statistics. Many of the commonly used parametric techniques have nonparametric equivalents, allowing for convenient use and application to traditional experimental and nonexperimental research designs and studies. Because nonparametric methods require less stringent distributional assumptions, these techniques provide statistical tests for analyzing data in conditions failing to satisfy the assumption of a normal ...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles